System and method of medical imaging

Abstract

A system of medical imaging including a marking module, a classifying module, a region-of-interest determining module, a curve acquiring module and a delay calculating module. The marking module is used for marking one or more feature regions from pre-scanned images; the classifying module is used for classifying the feature regions by using a classification algorithm model; the region-of-interest determining module is used for selecting a feature region of the same type as a specific diagnostic tissue as a region of interest; the curve acquiring module is used for acquiring a curve of CT values of the region of interest in function of time, based on a relationship between the CT values and scanning time points; the delay calculating module is used for detecting a peak value of the curve and calculating a scan delay time based on a time point corresponding to the peak value.

What is claimed is: 
     
       1. A system of medical imaging, comprising:
 a marking module for processing data, wherein the marking module is stored in a memory, marking feature regions from a plurality of pre-scanned images; 
 a classifying module for processing data, wherein the classifying module is stored in the memory, classifying the feature regions by using a classification algorithm model wherein the classification algorithm model comprises a K-means classification model or a Vector Space Model (VSM), and the classifying module uses feature parameters of the marked feature regions as input values of the K-means classification model or the VSM to acquire output values of the K-means classification model or the VSM, and classifies the feature regions based on the output values of the K-means classification model or the VSM and wherein the feature parameters of the feature regions comprise the following parameters: a relationship of CT values of pixels in the feature regions as a function of the scanning time points of the plurality of pre-scanned images, the number of pixels in the feature regions, 1st-Nth moments of the pixel values in the feature regions, and centroid of the pixels in the feature regions, where N is an integer more than 1; 
 a region-of-interest determining module for processing data, wherein the region-of-interest determining module is stored in the memory, selecting a feature region, among the classified feature regions, of a same type as a specific diagnostic tissue as a region of interest; 
 a curve acquiring module for processing data, wherein the curve acquiring module is stored in the memory, acquiring a curve of Computed Tomography (CT) values of the region of interest as a function of time, based on a relationship between the CT values of the region of interest in the plurality of pre-scanned images and scanning time points of the corresponding pre-scanned images; and 
 a delay calculating module for processing data, wherein the delay calculating module is stored in the memory:
 detecting a peak value of the curve; 
 calculating a scan delay time based on a time point corresponding to the peak value with regard to an origin of a time axis; and 
 determining, a transformed scan delay time based on at east one of: a composition, and a dosage of a contrast media. 
 
 
     
     
       2. The system of medical imaging according to  claim 1 , wherein the marking module comprises a fine division unit for judging whether a degree of linear correlation between individual pixels in the pre-scanned images and their surrounding pixels reaches a preset value, and marking a region where the degree of linear correlation of the pixels reaches the preset value as one of said feature regions. 
     
     
       3. The system of medical imaging according to  claim 1 , further comprising a mode match module for receiving a diagnostic mode set by an operator and removing a region not matching with the diagnostic mode from the determined regions of interest based on the diagnostic mode, wherein the mode match module is stored in the memory. 
     
     
       4. The system of medical imaging according to  claim 1 , further comprising a model modification module for receiving an invalid feature parameter and inputting the invalid feature parameter into the classification algorithm model to modify the classification algorithm model, wherein the model modification module is stored in the memory. 
     
     
       5. A method of medical imaging, comprising:
 marking feature regions from a plurality of pre-scanned images; 
 classifying the feature regions by using a classification algorithm model; wherein the classification algorithm model comprises a K-means classification model or a Vector Space Model (VSM), said classifying the feature regions by using a classification algorithm model comprising:
 using feature parameters of the marked feature regions as input values of the K-means classification model or the VSM to acquire output values of the K-means classification model or the VSM, and classifying the feature regions based on the output values of the K-means classification model or the VSM and wherein the feature parameters of the feature regions comprise the following parameters: a relationship of CT values of pixels in the feature regions as a function of the scanning time points of the plurality of pre-scanned images, the number of pixels in the feature regions, 1st-Nth moments of the pixel values in the feature regions, and centroid of the pixels in the feature regions, where N is an integer more than 1; 
 
 selecting a feature region, among the classified feature regions, of a same type as a specific diagnostic tissue as a region of interest; 
 acquiring a curve of Computed Tomography (CT) values of the region of interest as a function of time, based on a relationship between the CT values of the region of interest in the pre-scanned images and scanning time points of the corresponding pre-scanned images; 
 detecting a peak value of the curve; 
 calculating a scan delay time based on a time point corresponding to the peak value with regard to an origin of a time axis; and 
 determining a transformed scan delay time based on at least one of: a composition, and a dosage of a contrast media. 
 
     
     
       6. The method of medical imaging according to  claim 5 , wherein said marking feature regions from a plurality of pre-scanned images comprises fine division steps comprising:
 judging whether a degree of linear correlation between pixels in the pre-scanned images and their surrounding pixels reaches a preset value; and 
 marking a region where degree of linear correlation of the pixels reaches the preset value as one of said feature regions. 
 
     
     
       7. The method of medical imaging according to  claim 5 , further comprising, after said selecting a feature region of the same type as a specific diagnostic tissue as a region of interest, receiving a diagnostic mode set by an operator, and removing a region not matching with the diagnostic mode from the determined regions of interest based on the diagnostic mode. 
     
     
       8. The method of medical imaging according to  claim 5 , further comprising, after said selecting a feature region of the same type as a specific diagnostic tissue as a region of interest, receiving an invalid feature parameter, and inputting the invalid feature parameter into the classification algorithm model to modify the classification algorithm model.